Pages: [1]   Go Down

Author Topic: Gigapixel photos  (Read 12428 times)

didger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2030
Gigapixel photos
« on: May 18, 2005, 07:16:06 pm »

As you say, there's nothing very new about this except for packaging it into a single camera that's available to non-military buyers, but that's out of reach for almost every one and that surely no one could haul around in the field.  Slightly more feasible for the average photogapher is doing files this big by shooting huge arrays of frames and then very tediously putting them together with panorama stitching programs substantially customized to handle huge output like this.  

However, what I find more exciting is that I'm now able to do lots of images rivaling the resolution of 4x5 and even 8x10 film on a somewhat streamlined production basis without going to real extremes and without carrying a prohibitive amount of stuff in the field.  The stitch processing takes quite a long time even on a G5, and the really large images take several stages, but I'm doing other things on other computers in the meantime.  In the field I actually carry less than most photographers would carry without panorama stitch capability; a D2X, 3 lenses, an extremely light (about 2 lbs with ball head) Gitzo 1028, and a 5 oz. custom panorama bracket that stays on the camera all the time.  The stitching can now be done at very high quality and very modest price (though considerable learning curve) for PC (with PTAssembler) and Mac (PTMac).
Logged

Bobtrips

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 679
Gigapixel photos
« Reply #1 on: May 19, 2005, 12:35:59 pm »

Quote
So my imaginary ultra-high resolution camera uses a high resolution sensor (the smallest viable pixels), an extremely sharp lens (macro?, telephoto?), a very solid tripod, and a robotic camera pointing and control system to automate taking the multiple frames.
Why not multiple cameras in shooting in 'ganged' format?

Say a dozen Oly C8080s - 8 megs * 12 (minus a bit for overlap).

The 'multi' would be limited to ISO 100 and lower.  Hardly a major problem when shooting off a tripod.

With all cameras firing at once the moving grass/breaking wave problems disappear.  (And the price would be very attractive when compared to MF digital or scanning backs.)

I'm not sure about the nodal point issues.  Does this kill the idea for shots because of alignment problems?
Logged

Bobtrips

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 679
Gigapixel photos
« Reply #2 on: May 19, 2005, 03:15:59 pm »

Quote
You also have the additional problem of getting the identical color and luminance response from an array of consumer-grade sensors, which adds another difficult obstacle that must be overcome to stitch successfully.
That would seem to be an easy software fix.

Shoot a standardized target and let the software determine 'fixes' for each camera.

Stitching alignment should be easy as well.  Again software.  Each camera should be easily adjustable and software can report on degree of overlap.
Logged

didger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2030
Gigapixel photos
« Reply #3 on: May 19, 2005, 04:29:08 pm »

Is this stuff supposed to be serious, or a just for fun weird disscussion?  I'm just about OK with carrying a D2X, a 2 lb. tripod (complete with ball head), 3 lenses, and a permanently mounted 5 oz. panorama bracket and a spare 3.5 oz battery and a few CF cards.

Is the ANYBODY that would want to carry a nightmare monster camera array and ultra nightmare monster tripod to hold this, and then spend an hour or two getting it all set up and aligned for each shot?  You're all just kidding, right?  Right?
Logged

Bobtrips

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 679
Gigapixel photos
« Reply #4 on: May 19, 2005, 05:18:16 pm »

Well, for me this is an intellectual exercise.  I have no intention of lugging a massive D2x around with me.  

The Oly C8080 weighs 25.5 oz.  Let's downscale the camera for sake of the discussion.

The Casio EX-Z750 weighs 5.8 oz., has a 7 meg sensor and a 3x lens.  For the sake of discussion let's not get into the qualities of this particular camera or whether it can be remotely controlled.  (Just play along for fun.  Or go to another sand box if you don't like this game.)

That would mean that a dozen of them would weigh 69.6 oz. or 4.35 lbs.  Add in a bit more for a carbon fiber 'rack'.  And a couple of pounds for a sub-notebook to run the show.

Now you've got a 'less than ten pound' 84 meg camera (minus a few pixels for overlap).

The cameras are going to cost you less than $5k.  The rack?  And sub-notebook.

And a well designed rack should be rapidly reconfigured for 3x4, 2x6, 1x12 shots.  That gives one a range of something close to square to ultra-pano.

And hour or two to set up?  That part I seriously doubt.  Cameras mounted to rack and a carbon fiber/titanium clamshell case with shoulder straps.  

Alignment with software/computer assistance should take only moments (assuming that the individual camera mounts are well designed).

Is there a market?  I have no way of knowing.  But if LF film is equal to about 70 megs and people find a need for LF film....
Logged

Bobtrips

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 679
Gigapixel photos
« Reply #5 on: May 19, 2005, 07:32:49 pm »

The market?  Again, I have no idea.

Let's look at stitching for a moment.  How good is it at scenes with lots of movement?  Say you get hired to make a very large print of the horses crossing the finish line at Churchill Downs?  

Now I don't know Jack about LF film.  So I went to B&H's web page and found an 8 x 10 for a bit over $5k.  That's basic camera only.  No lens, no whatever else you need to make it fly.  And it weighs about 20 pounds.

I do remember looking up the cost of film for a 8 x 10 at one point and seem to remember that it was in terms of dollars per shot.

I don't think I'll bother calling Walgreen's to see if they can give me one hour turnaround on my prints from the beast.

So if one could cut 50% off the weight.  (Probably more if the cameras components were mounted without all the extra batteries, viewfinders, LCD screens, and doodahs.)

And if one could get the digital advantages of no-cost clicking, instant digital format, etc.

And if one weren't locked into a single dimension rectangle.

?
Logged

DiaAzul

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 777
    • http://photo.tanzo.org/
Gigapixel photos
« Reply #6 on: May 19, 2005, 08:58:21 pm »

Quote
Say you get hired to make a very large print of the horses crossing the finish line at Churchill Downs?  
That's been going on for years and the technology has been in place for a long time - it's what is traditionally what is looked at in a photo-finish.

In these situations a roll of film is fed passed a slot. The background will be blurred, but you will have a long (3-4ft) picture of all the horses passing the line. Don't know how many mega-pixels this equates to, and what the relevance is to zigapixel cameras. But what the heck.

I came across a scanning camera which does high dynamic range, high mega-pixels and 360 degree panoramas - is very light, automatic and looked really neat and cool. Problem is I can't find it again (it is manufactured by some German company).

This is something else I came across while trying to find the German cam...Pano-cam

Found the german cam - they do both panorama, spherical and HDR versions for 360 degree images:

Spheron
Logged
David Plummer    http://photo.tanzo.org/

didger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2030
Gigapixel photos
« Reply #7 on: May 20, 2005, 01:57:19 am »

There's probably several existing solutions and unlimited theoretical solutions for that one special hypothetical problem of a client that wants to make a HUGE print of a horse race.  Just how often would such a situation arise, where an ordinary 4x5 camera would be inadequate?  Lots and lots of commercial photographers have 4x5 cameras, so unless your client needs really really HUGE, you would not be in a particularly favorable position to get the gig with the big array camera and I doubt that anyone would ever develop such a kludge just for that tiny (perhaps altogether non-existent) niche market, especially since (as Dia mentioned) there are already other solutions that would work in those very rare and special situations.

Morever, I very seriously doubt that the camera array and adequate physical support and field laptop and adequate power supply would be so light.  I doubt this even more seriously if this thing also has to be easily configurable for various array modes (3x4, 6x2, 12x1).  The precision to make this possible to do quickly and with perfect alignment would not be easy and certainly not very light.  If I were looking to invest some money for developing some sort of new gadget, this would not be the one.  On the other hand, I think larger and larger sensors are inevitable and someone with a viable way to produce them cheaper would be a more tempting investment opportunity.

Thin air conceptualizing is SO easy.  Wanna go to the moon?  Just make a powerful rocket and appropriate control and guidance and safety and life support systems and you're on your way.  Implementation is just engineering details.
Logged

Image Northwest

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 63
Gigapixel photos
« Reply #8 on: May 18, 2005, 06:48:25 pm »

For those of you who really want the ultimate, there's a camera out there that doesn't measure its file size in megabytes but in gigabytes.  This was mentioned the recent CPU magazine, briefly explaining the process and potential.  The current working camera produces a 4 gigabyte file.  The technology has been around for a while, mostly under military scrutiny, using parts designed for spy cameras and components from 9 x 18 inch reconnaissance equipment.  Actually the process is not new.  The image is recorded on film, then scanned and worked through PS. Something we all do, or have done.

Obviously the camera is fairly sizable, not one to tote without mechanial help, but the pictures are, needless to say, incredible and of such high-resolution that the photo can reveal detail not readily visible to the human eye.    Not a camera priced for the average or even high end consumer, but interesting nevetheless.  For more information go to:

www.gigapxl.org
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Gigapixel photos
« Reply #9 on: May 19, 2005, 10:00:34 am »

The "gigapxl" project and its website http://www.gigapxl.com are interesting, with lots of carefully described technical details about the challenges of ultra-detailed imaging. I have one significant quibble though, which means that for practical purposes, they really have "only" about 100 million useful pixels worth of detail.

This can easily be seen in some of the crops provided; despite corresponding to about a megapixel of the image or more, they are as soft and lacking in fine detail as web-formatted images of less than 100,000 pixels.

The problem come from the fact that at each of the numerous stages that limit resolution (amospheric turbulelence, lens abberations, diffraction, film resolution, etc.) they require 50% MTF, a fairly standard design target in optics. In plain language, they require no more than a one-stop loss of contrast at each stage. But these losses combine; the percentages multiply. So with about five such stages, their system loses about five stops of contrast, or has combined MTF of around 5%, and this is generally considered to be so low as to be useless, unless your goal is to photograph ultra-high contrast transmissive test patterns.

One thing that interests me a lot is their study of atmospheric effects; they need to use an ultrawide FOV to get atmospheric limits under control; for a normal angular field of view, 50% MTF is limited to about  about 3,000 line pairs per picture height. With a medium that dos not use Bayer inerpooltion (like ascanning back), that gives a useful limit of 6,000 pixels high. With Bayer pattern sensors, maybe up to 10,000 pixels high.

Adding in other unavoidable resolution losses due to lenses, and the current 4,000 pixel high 22MP sensors are about half way to the useful limit as far as "line pairs of resolution", or a quarter of the way if you count pixels.

Scanning backs already go beyond 6,000 pixels high, so they are probably already at the limit for angular FOV out to moderately wide, with only ultra-wide likely to benefit from much more.


But that brings me to didger's suggestion of blending frames.

Once the sensor is scanning rather than single shot, so that you need a completely stationary subject, it is probably easier and better to blend multiple frames. For one thing, bleding is much easier on lens design, since you can use a lens of normal to narrow FOV, for which optical abberations are far easier to control. For another, one can add in dynamic range blending with mutliple frames on each part of the scene.

So my imaginary ultra-high resolution camera uses a high resolution sensor (the smallest viable pixels), an extremely sharp lens (macro?, telephoto?), a very solid tripod, and a robotic camera pointing and control system to automate taking the multiple frames.
Logged

Jonathan Wienke

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5829
    • http://visual-vacations.com/
Gigapixel photos
« Reply #10 on: May 19, 2005, 03:10:13 pm »

Quote
I'm not sure about the nodal point issues.  Does this kill the idea for shots because of alignment problems?
Yep. If the frames are properly aligned, you can deal with subject motion in most cases by creatively choosing the blend boundary between frames, but if the alignment is off (as it would have to be with ganged cameras) blending is MUCH harder. You also have the additional problem of getting the identical color and luminance response from an array of consumer-grade sensors, which adds another difficult obstacle that must be overcome to stitch successfully. One-camera pan & scan is much more practical.
Logged

Jonathan Wienke

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5829
    • http://visual-vacations.com/
Gigapixel photos
« Reply #11 on: May 19, 2005, 04:12:59 pm »

Quote
Shoot a standardized target and let the software determine 'fixes' for each camera.
I can see that working reasonably well.

Quote
Stitching alignment should be easy as well.  Again software.  Each camera should be easily adjustable and software can report on degree of overlap.

Ain't so. Even if you're shooting with one camera from a tripod, if the nodal point adjustment is off, stitching properly becomes nearly impossible if there are foreground elements. I have some shots I took from Bright Angel Point at the north rim of the Grand Canyon, where the lower foreground is the rock I'm standing on (~4-20 feet from the lens), and the background is the canyon around me (all at infinity focus). I screwed up the nodal point setting, and the shots will NOT stitch decently. I always get something resembling a constipated snake, because the foreground offset due to the nodal misadjustment totally befuddles the stitching program, even with a 50% frame-to-frame overlap. Shooting with an array of lenses physically offset would be an even worse mess to unsnarl.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Gigapixel photos
« Reply #12 on: May 19, 2005, 04:43:31 pm »

Quote
Is the ANYBODY that would want to carry a nightmare monster camera array and ultra nightmare monster tripod to hold this, and then spend an hour or two getting it all set up and aligned for each shot?

You're all just kidding, right?  Right?
To question 1: "yes, but not me"; the same sort of folks who haul 8"x10" view cameras up mountains and down slot canyons.

To question 2: I am certanly not: I am trying to find something sufficiently difficult to be spiritually satisfying for those image quality obsessives who deeply believe that "if the process is not hugely time consuming and painful, the results cannot possibly be good enough".
Logged

didger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2030
Gigapixel photos
« Reply #13 on: May 19, 2005, 05:10:07 pm »

Ah, I see (sort of, I suppose).
Michael Fatali is the only guy I know that shoots 8x10 on mountains and in slot canyons and he avoids doing the hauling himself.  He has adoring young accolytes (college kids) that do the hauling for him in the hopes that some of the magic rubs off on them (or something).

Hey, BJL, maybe some time you could find solutions for even greater weight savings and improved image quality for those of us who find "spiritually satisfying" photography in liberation from excessive weight and encumbrances.  Those other guys can always handicap themselves with lead weightbelts like divers use, right?
Logged

didger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2030
Gigapixel photos
« Reply #14 on: May 19, 2005, 06:34:57 pm »

Well, it's an interesting sandbox, but I'll go no farther down this road than fun discussion.  84 megs under 10 lbs?  So what?  I've got hundreds of megs (if I want) for under 8 lbs. and with all sorts of versatitilty and alternative options (like normal digital photography with normal single files, more varied framing with various zoom lenses, hand held if I want, etc.)  What are the advantages of this ultra-specialized array configuration at the high cost that always comes with anything ultra-specialized compared to fast and efficient and dirt cheap panorama shooting with a single camera?  You can't compare this array contraption plus laptop plus special rack with an 8x10 camera.  There's hardly anything simpler than an 8x10 camera and even so, how big is the market for that?
Logged

didger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2030
Gigapixel photos
« Reply #15 on: May 19, 2005, 08:03:32 pm »

Well, I don't see this thing as likely ever happening (compared to multi-component single sensor single lens large format digital) and I doubt exceedingly that the price would be anywhere near as low as a large format kit.  However, I was dead wrong in my predictions about the unlikelihood of a high density small format sensor with high quality small format lenses to match it.  Owning a D2X is a sincere gesture of admitting that I was wrong.  I could be wrong about this array thing too.  The reason I stubbornly cling to not believing in it is not for any inherent technical problems, but simply because the market would only be for those folks now buying 8x10 cameras, and that's a VERY small market, so the development cost would have to be covered by very few customers and I also doubt very much that the development would be as quick and easy as our airy discussions might lead us to believe.  Remember Pournelle's Law:  It always takes longer and costs more.
Logged

Image Northwest

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 63
Gigapixel photos
« Reply #16 on: May 19, 2005, 10:51:33 pm »

It's amazing the direction of this thread.  I mentioned the gigabyte camera only because I thought it was an interesting method for someone who is specializing in a a very specific product.  My own camera methods are most like Didger's, especially in my attempt to get the best photo possible with a reasonable weight, since I too do lots of backpacking (and long day trips). So I would never succumb to the gigabyte format.  However, for some people it might work.  The most amazing thing about today's photography spectrum is the variations that are possible and the choices available in our attempt to get the "picture".  Let evolution continue.

Bruce
Logged

gwarrellow

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 65
Gigapixel photos
« Reply #17 on: May 20, 2005, 05:42:15 am »

Quote from: DiaAzul,May 19 2005,20:58
Quote from: Bobtrips,May 19 2005,19:32
I came across a scanning camera which does high dynamic range, high mega-pixels and 360 degree panoramas - is very light, automatic and looked really neat and cool. Problem is I can't find it again (it is manufactured by some German company).

This is something else I came across while trying to find the German cam...Pano-cam

Found the german cam - they do both panorama, spherical and HDR versions for 360 degree images:

Spheron
Thanks for the links David.

I came across a Swiss company by the name of Seitz that has been in the panoramic photography business for over 50 years.  They produce a range of film and digital based solutions both for "35mm" and medium format:

http://www.roundshot.ch/xml_1/internet/de/...cation/f469.cfm

The quality of the products are excellent - I recently purchased the Roundshot VR Drive for my 1Ds.  This unit sits on top of my tripod and can be programmed to take set panoramic angles, bracketed shots at each angle, set rotation time between shots, etc.  The unit connects via a supplied cable directly to the camera.  Very cool!

http://www.roundshot.ch/xml_1/internet/de/...n/d438/f629.cfm

I thought the bracketed shot option might be very cool when used in combination with Photoshop CS2's new 23-bit HDR function.

They were VERY professional to deal with and are multilingual.  They even shipped the unit to me free of charge.  HIghly recommended.

Graham
Logged
Pages: [1]   Go Up